![Arxiv:0712.0004V1 [Astro-Ph] 30 Nov 2007 Erhcuclo Aaa H Etentoa El Ehrh Sc Recherche La De National Centre Hawaii](http://data.docslib.org/img/6f676759ed6a1885aa4608519dc246de-1.webp)
arXiv:0712.0004v1 [astro-ph] 30 Nov 2007 erhCuclo aaa h eteNtoa el ehrh Sc Recherche la de National Centre Hawaii. the Canada, of Council search colo hsc n srnm,Uiest fS nrw,N Andrews, St of University Astronomy, and Physics of School 1 ae nosrain olce tteCnd-rneHwi Te Canada-France-Hawaii the at collected observations on Based ihrslto ehlesetacletdwt SaOSoe few a th over Jupiter In ESPaDOnS hot with with stars orbit. collected 7 planet’s ´echelle spectra of high-resolution the activity chromospheric to the synchronized monitored we activity stellar of variation t o uie ( Jupiter hot its ezegIsiuefrAtohsc,Ntoa eerhCo Research National Astrophysics, for Institute Herzberg vdnesgetn nosral antcitrcinbtena between interaction magnetic observable an suggesting Evidence h nO aueo trPae Interactions Star-Planet of Nature On/Off The AAAtoilg nttt,Uiest fHwi tManoa at Hawaii of University Institute, Astrobiology NASA P orb 24HwetPae itra C 8 4P7 V8S BC, Victoria, Place, Hewlett 1234 60Wolw rv,Hnll,H 96822 HI Honolulu, Drive, Woodlawn 2680 < [email protected] days, 7 itraB,Cnd 9 2E7 V9E Canada BC, Victoria [email protected] [email protected] nrwClirCameron Collier Andrew odnAH Walker A.H. Gordon ai .Bohlender A. David [email protected] < a vey Shkolnik Evgenya ieK1 9SS KY16 Fife ABSTRACT . AU, 0.1 and M p sin > i etfiu fFac,adteUiest of University the and France, of ientifique . M 0.2 ecp prtdb h ainlRe- National the by operated lescope J rhHuh tAndrews, St Haugh, orth per sacyclic a as appears ) ni fCanada of uncil sn new using s ihsin nights sstudy, is trand star 1 –2–
2005 and 2006 from the CFHT. We searched for variability in several stellar activity indicators (Ca II H λ3968, K λ3933, the Ca II infrared triplet (IRT) λ8662 line, Hα λ6563 and He I λ5876). HD 179949 has been observed almost every year since 2001. Synchronicity of the Ca II H & K emission with the orbit
is clearly seen in four out of six epochs, while rotational modulation with Prot=7 days is apparent in the other two seasons. We observe a similar phenomenon
on υ And, which displays rotational modulation (Prot=12 days) in September 2005, in 2002 and 2003 variations appear to correlate with the planet’s orbital period. This on/off nature of star-planet interaction (SPI) in the two systems is likely a function of the changing stellar magnetic field structure throughout its activity cycle. Variability in the transiting system HD 189733 is likely associated with an active region rotating with the star, however, the flaring in excess of the rotational modulation may be associated with its hot Jupiter. As for HD 179949, the peak variability as measured by the mean absolute deviation (MAD) for both HD 189733 and τ Boo leads the sub-planetary longitude by ∼ 70◦. The
tentative correlation between this activity and the ratio of Mpsini to the planet’s rotation period, a quantity proportional to the hot Jupiter’s magnetic moment, first presented in Shkolnik et al. 2005 remains viable. This work furthers the characterization of SPI, improving its potential as a probe of extrasolar planetary magnetic fields.
Subject headings: stars: late-type, activity, chromospheres, planetary systems, radiation mechanism: non-thermal, stars: individual: τ Boo, HD 179949, HD 209458, HD 189733, HD 217107, HD 149143
1. Introduction
Observations and theory demonstrate that star-planet interaction (SPI) is a complex, yet potentially very informative probe of extrasolar planetary magnetic fields. In Shkolnik et al. (2003, 2005a), we reported on planet-induced chromospheric activity on two stars, HD 179949 and υ And apparent from the night-to-night modulation of the Ca II H & K chromospheric emission phased with the hot Jupiter’s orbit. The modulation was indicative of a magnetic rather than tidal interaction (Cuntz et al. 2000), such that the period of the observed stellar activity correlated with the planet’s orbital period Porb, rather than Porb/2. Ample observational evidence of tidal and magnetic interactions exists in the exaggerated case of the RS Canum Venaticorum (RS CVn) stars, which are tightly-orbiting binary sys- tems consisting of two chromospherically active late-type stars (e.g. Glebocki et al. 1986, –3–
Catalano et al. 1996, Shkolnik et al. 2005b ). Although efforts to observe variable radio emission from the stars with hot Jupiters have not yet been successful (e.g. Lazio & Farrell 2007, George & Stevens 2007), there have been several additional observations that support the existence of SPI. Photometric observations by the MOST space telescope of several hot-Jupiter systems, including HD 179949 and τ Boo, suggest that stellar surface activity in the form of active spots may be induced by the giant planet (Walker et al. 2006, 2007). Also, Saar et al. (2006) recently reported a possible detection of planet-induced X-ray emission from the HD 179949 system corresponding to ∼30% increase in X-ray flux over quiescent levels coincident with the phase of the Ca II
enhancements at φorb∼0.8. A statistical analysis by Kashyap et al. (2006) suggests that the X-ray flux from stars with hot Jupiters is on average & 3 times greater than stars with planets at larger orbital distances, presenting further evidence that close-in giant planets have a measurable effect on the activity of the parent star. One scenario of magnetospheric interaction proposes that the planet induces reconnec- tion events as it travels through the large stellar magnetic loops (Cuntz et al. 2000, Ip et al. 2004), implying that the resulting activity should depend on the star’s magnetic field, the planet’s magnetic field and the orbital distance with respect to the Alfv´en radius of the host star (∼10 stellar radii). Novel research of the magnetic field topology of hot Jupiter host stars is underway (Catala et al. 2007, Moutou et al. 2007) using Zeeman Doppler Imaging (ZDI), which hopes to contribute to a more detailed understanding of SPI. A detection of a magnetic field of a hot Jupiter would 1) provide a constraint on the rapid hydrodynamic escape of its atmosphere (Vidal-Madjar et al. 2003, 2004) which could affect the planet’s structure and evolution, 2) present implications for the planet’s internal structure, and 3) shed light on the mass-radius relationship of the known transiting planets (Pont et al. 2005, Bakos et al. 2006). Although the internal magnetic fields of hot Jupiters are expected to be weaker than Jupiter’s due to probable tidal locking and slower spin rates (Sanchez-Lavega 2004, Griessmeier et al. 2004), Olson & Christensen (2006) calculated that the magnetic field of a planet with even a tenth of Jupiter’s rotation rate would still have a strong dipole moment, when reasonably assuming that the convection is not highly modified by the rotation rate. Also, the fact that both hot and very hot Jupiters, such as HD 209458 b and OGLE-TR-56 b, are detected at all means that they must have strong enough magnetic fields to balance the extreme stellar irradiation and CME plasma pressure to prevent destructive atmospheric erosion (Khodachenko et al. 2007). We seek to probe hot Jupiter magnetic fields in order to understand their formation and evolution. SPI potentially offers an indirect way to detect, and with future modeling and observations, measure planetary magnetic fields. –4–
It is reasonable to assume that any magnetic interaction would be greatest in the out- ermost layers of the star, namely the chromosphere, transition region and the corona due to their proximity to the planet, low density, and non-radiative heat sources. With the commis- sioning of CFHT’s high-resolution ´echelle spectrograph, ESPaDoNS, we are able to include several stellar activity indicators in our analysis to observe the interaction as a function of atmospheric height in order to model the energy transfer and dissipation mechanisms of this phenomenon. ESPaDOnS’ wavelength coverage allows simultaneous monitoring of the Ca II infrared triplet (IRT, lower chromosphere), Hα, Ca II H, K (middle chromosphere), and He ID3 (upper chromosphere). Our program stars have planets with orbital periods between 2.2 and 7.1 days, eccentric- ities ≈ 0 and semi-major axes < 0.08 AU. These systems offer the best chance of observing upper atmospheric heating. Of the seven systems we observed with ESPaDoNS, τ Boo, HD 179949, υ And and HD 209458 have been observed previously in our CFHT/Gecko campaign. The first results from 2001 and 2002 observations, including the first evidence of planet-induced magnetic heating of HD 179949, were published in Shkolnik et al. (2003, 2005a). We later extended the experiment at the Very Large Telescope (VLT) to include five southern targets. The three new systems monitored in this study are HD 217107, HD 149143 and HD 189733. The system parameters for the ESPaDOnS program stars are listed in Table 1 along with our standard 61 Vir. In this paper, we present new ´echelle spectra and compare with those of previous years, bringing to light a broader understanding of stellar activity, its cycles, and SPI. The details of our observations and data reduction are outlined in Section 2. In Section 3 we discuss our analysis and results of the Ca II K measurements including long-term, short-term and rotational modulation. Comparisons with other activity indicators are made in Section 4. –5–
Table 1. Stellar and Orbital Parameters
a a a b ′ c d Star SpT vsini Prot Porb Mpsini a hKi hK i hMADKi He I EW −1 (km s ) (days) (days) (MJ ) (AU) (A)˚ (A)˚ (A)˚ (mA)˚
τ Boo F7IV 14.8±0.3 3.2e 3.31 4.4 0.046 0.336 0.184 0.0019 27 HD179949 F8V 6.3±0.9 7f 3.09 0.98 0.045 0.369 0.186 0.0022j 17 HD209458 G0V 4.2±0.5 16g 3.53 0.69h 0.045 0.195 0.078 0.0009 .3 υ And F7 V 9.0±0.4 12e,k 4.618 0.71 0.059 0.254 0.091 0.0016 – HD189733 K1V 2.92±0.22 11.7i 2.22 1.15h 0.031 1.337 1.231 0.0044j 35 HD217107 G8IV 9.0±0.4 39k 7.13 1.35 0.075 0.160 0.075 0.0007 <4 HD149143 G0IV 3.9±1 ?? 4.09 1.36 0.052r 0.342 0.144 0.0014 10 61 Vir G5 V 2.2m 33k — — — 0.182 0.083 0.0008 <4
aPublished orbital solutions: τ Boo − Butler et al. 1997, HD 179949 − Tinney et al. 2001, HD 209458 − Charbonneau et al. 1999, υ And − Butler et al. 1997, HD 189733 − Bouchy et al. 2005, HD 217107 − Fischer et al. 1998, HD 149143 − De Silva et al. 2006 bTotal integrated intensity of the mean normalized Ca II K core. These values are relative to the normalization points near 3930 and 3937 A˚ at 1/3 of the pseudo-continuum at 3950 A.˚ cWe subtracted the photospheric emission from hKi in order to measure the mean integrated chromospheric emission hK′i using data from Wright et al. (2004). (See text for more details.) dAverage integrated ‘intensity’ of the mean absolute deviation (MAD) of the K residuals, per observing run eHenry et al. 2000 f This work and Wolf & Harmanec 2004 gMazeh et al. 2000 hTransiting system iCroll et al. 2007 jValue were corrected to remove geometric (rotational and/or planetary) modulation of an active region on the star. For HD 189733, the non-corrected value is 0.0098 and for HD 179949, 0.0063. kWright et al. 2004 –6–
Table 2. Observations
Star U B V Exposuresa S/N b S/N b t × n × N at 3950 A˚ at8710 A˚
τ Boo 5.02 4.98 4.50 120 × 10 × 9c 630 1640 HD 179949 6.83 6.76 6.25 660 × 5 × 7e 440 1270 HD 209458 8.38 8.18 7.65 1800 × 4 × 4 350 960 υ And 4.69 4.63 4.09 180 × 120 × 4e 2800 7290 HD 189733 − 8.60 7.67 1800 × 4 × 4 280 1260 HD 217107 7.33 6.90 6.18 600 × 5 × 3 350 1340 HD 149143 − 8.53 7.90 1800 × 4 × 4 320 1000 61 Vir 5.71 5.45 4.74 120 × 7 × 5 450 1480
at = Exposure time in seconds, n = number of exposures per night, N = number of nights during the June 2006 ESPaDOnS observing run bTypical nightly S/N per 0.022-A˚ pixel cSpectra from three of the nine nights were observed in ESPaDOnS’ ‘spectropo- larimetry’ mode. dFour nights in June 2006 and three nights in September 2005 eAll these data were acquired in the ‘spectropolarimetry’ mode in September 2005. The extremely high S/N was a requirement of the partner program to search for linear polarization. (Collier Cameron et al., in prep.) –7–
2. The spectra
The observations were made with the 3.6-m Canada-France-Hawaii Telescope (CFHT) on 7 nights in September 2005 and 9 nights in 2006 June. We used ESPaDOnS (Echelle´ Spec- troPolarimetric Device for the Observation of Stars), which is fiber fed from the Cassegrain to coud´efocus where the fiber image is projected onto a Bowen-Walraven slicer at the spectrograph entrance. With a 79 gr/mm grating and a 2048×4608-pixel CCD detector, ESPaDOnS’ ‘star-only’ mode records the full spectrum over 40 grating orders covering 3700 to 10400 A˚ at a spectral resolution R of ≈80,000. The four nights of observations of υ And (18, 19, 23, 24 September 2005) and three nights of the τ Boo data (16, 17, 18 June 2006) were taken in ESPaDOnS’ ‘spectropolarimetry’ mode at R of 68,000. The data were reduced using Libre Esprit, a fully automated reduction package provided for the instrument and described in detail by Donati et al. (1997, 20071). Each stellar exposure is bias-subtracted and flat-fielded for pixel-to-pixel sensitivity variations. After optimal extraction, the 1-D spectra are wavelength calibrated with several Th/Ar arcs taken throughout the night. Finally the spectra are divided by a flat-field response and then the continuum is normalized. Heliocentric velocity corrections are applied as well as small velocity corrections (< 100 m s−1) to account for instrumental effects using the telluric lines. The final spectra were of high S/N reaching ≈ 130 per pixel (880 A˚−1) in the H & K emission core, 400 per pixel (2700 A˚−1) in the pseudo-continuum near 3950 A,˚ and about 3 times higher near the Ca II IRT. Spectra with comparable S/N were taken of 61 Vir, a G5 V star known not to have close-in giant planets, plus the hot standard HR 5511 (SpT = A0V) for telluric line correction. Table 2 lists the program stars, including their magnitudes, exposure times, and typical S/N. All further processing and analysis were performed with standard IRAF (Image Reduc- tion and Analysis Facility) routines.1 Differential radial velocity corrections were applied to each stellar spectrum using IRAF’s fxcor and rvcorrect routines. Representative spectra near the key stellar activity indicators are shown in Figure 1 for HD 189733.
1Also see http://www.cfht.hawaii.edu/Instruments/Spectroscopy/Espadons/Espadons esprit.html. 1IRAF is distributed by the National Optical Astronomy Observatories, which is operated by the As- sociation of Universities for Research in Astronomy, Inc. (AURA) under cooperative agreement with the National Science Foundation. –8–
3. Measuring chromospheric activity
The very strong Ca II H and K photospheric absorption lines suppress the local stellar continuum making it difficult to normalize each spectrum consistently. The normalization level was set at 0.3 of the flux at 3950 A˚ centered on the H and K lines. Therefore the wavelengths were constant for all spectra of a given star, though they varied slightly from star to star due to variations in spectral type. The ≈7-A˚ spectral range was chosen to isolate the H and K reversals. This window is wide enough that a few photospheric absorption features appear to test for general stability. To normalize each sub-spectrum, the end points were set to 1 and fitted with a straight line. The mean Ca II K cores for four of the program stars are shown in Figure 2 with those for HD 179949, HD 189733 and υ And in Figures 3 − 5. The spectra were grouped by date and a nightly mean was computed for each of the lines. Other than the active HD 189733 (see Section 3.2), all stars observed had non-varying K emission at the . 0.001 level on average on a given night. This is a result of S/N variations and intranight (short time-scale) chromospheric activity. We used nightly residuals from the average stellar spectrum to measure the chromo- spheric activity within the reversals. Each residual spectrum had a broad, low-order curva- ture removed. The residuals of the normalized spectra (smoothed by 17 pixels) were used −1 to compute the mean absolute deviation (MAD = N Σ|datai − mean| for N spectra), a measure of overall variability within the span of the observing run. The Ca II K MAD spectrum and the nightly residuals used to generate it for HD 179949, HD 189733 and υ And are displayed in Figures 3 − 5. The analysis presented in this section consists of only Ca II K emission measurements for several reasons: 1) The broad, deep photospheric absorption of the Ca II K line allows the chromospheric emission to be seen at higher contrast as compared to Hα and the Ca II IRT where chromospheric emission merely fills in the absorption core. 2) Extensive studies by the Mount Wilson group and Wright et al. (2004) allow us to isolate the average chromospheric emission hK′i by correcting for the photospheric contribution to our measurements (hK′i =
hKi−hKphoti, see Section 3.3.2 of Shkolnik et al. 2005a for more details.) This makes for a more accurate comparison of the stars in the sample whose spectral types vary from F7 to K1. 3) Previous CFHT/Gecko spectra consisted of only a single order containing Ca II H & K and it is useful to make comparisons between the data sets. 4) Lastly, there are no telluric features or blended lines to contaminate the spectra, as is the case for the other indicators, which are discussed in Section 4. –9–
3.1. HD 179949 & υ And: Evidence of the on/off nature of SPI
When monitoring chromospheric emission, stellar activity may be modulated by the star’s rotation, planetary motion in the case of SPI, or a combination of both. The orbital periods of the planets are well known and uniquely established by the PRV and transit discovery methods, but the rotation periods of the stars are much harder to determine in part due to stellar differential rotation. For studies of SPI, differentiating between rotational and orbital modulation of the chromospheric emission is key. In Shkolnik et al. (2005a) we presented evidence of planet-induced heating on HD 179949. The effect lasted for over a year and peaked only once per orbit, suggesting a magnetic interaction. In the simplest configuration, a magnetic interaction would occur near the sub- planetary point, when the planet is in front of the star relative to the line-of-sight, which defines orbital phase φorb = 0. Reproduced in Figure 6, we fitted a truncated, best-fit spot model to our 2001 and 2002 data with P = Porb = 3.092 d corresponding to the change in projected area of a bright spot on the stellar surface before being occulted by the stellar limb. The fit to the 2001 and 2002 data peaks at φorb = 0.83 ±0.04 with an amplitude of 0.027. We over-plot new data from 2005 which is fit remarkably well by the same model with only an insignificantly small relative phase shift of -0.07. This phase lead may help identify the nature of the interaction. For example, the offset from the sub-planetary point of a starspot or group of starspots can be a characteristic effect of tidal friction, magnetic drag or reconnection with off-center stellar magnetic field lines. For further discussion on such mechanisms, see papers by Gu et al. (2005), Preusse et al. (2006) and McIvor et al. (2006). In any case, the phasing, amplitude and period of the activity have persisted for over 4 years. Ca II data acquired in 2003 and 2006 of HD 179949 do not phase with the planet’s orbit (Figure 7), but both phase well with a 7-day period, likely the rotation period of the star. In Figure 8 we fit data from each year separately with a rotation curve because the effects of differential rotation and the appearance and disappearance of new spots over the three years would produce variations in phase, amplitude and period in the observed modulation. Note that the amplitude of the rotational activity in only 0.6 of that of induced by SPI.
Indirect indications of the rotation rate of HD 179949 imply Prot ≈ 9 days and are presented in Shkolnik et al. (2003) and Saar et al. (2004). Wolf & Harmanec (2004) weakly detect (1.5 σ) a photometric rotation period for HD 179949 of 7.07 d with an amplitude of only 0.008 mag. While more photometry is needed to determine a rotation period conclusively, the modulated Ca II emission of this star in both 2003 and 2006 strongly suggests a rotation period of 7 days. – 10 –
Similarly, previous Ca II data of υ And indicated possible SPI (Figure 8 of Shkolnik et al. 2005a), yet our September 2005 data appears to vary with the rotation. Again, the rotation period is not well known. Henry et al. (2000) quotes both 11 and 19 days, with a probable 11.6-day period from the hSHKi index. We plot the 2005 data against the 4.6-d orbital in Figure 9 and an 11.6-day rotation period in Figure 10. Unlike data from 2002 and
2003, the 2005 data phase much better with Prot=11.6 d than with a planetary orbit. This on/off characteristic of SPI observed in the HD 179949 and υ And systems is predicted by the models of Cranmer & Saar (2007). They model the Ca II H & K light curve of a sun-like star with a hot Jupiter interacting with the field geometry at various stages of the empirically derived solar magnetic field at annual steps of the 11-year solar cycle. They conclude that due to the complex nature of the multipole fields, the Ca II K light curves due to SPI do not repeat exactly from orbit to orbit, and at times the planet-induced enhancement may disappear altogether leaving only rotationally modulated emission. This may explain the 2003 and 2006 disappearance of the strong orbital modulation seen in 2001, 2002 and 2005 for HD 179949. Their models also show that for sparsely sampled data, the apparent phase shift between the peak Ca II emission and the sub-planetary point may fall between 0.2 and +0.2 (or ± 72◦), consistent with the -0.17 phase shift we detected repeatedly for HD 179949.
3.2. HD 189733: The active host of a massive planet
We reported in Shkolnik et al. 2005a that chromospheric variability of the active, young star κ1 Ceti (hK′i=0.815), for which the presence of a hot Jupiter is not ruled out, and
HD 73256, known to host a hot Jupiter (Mpsini=1.85 MJ , a=0.037 AU, Udry et al. 2003, hK′i=0.899) was modulated by stellar rotation with additional variability or flaring poten- tially induced by a hot Jupiter. We find a similar effect on HD 189733, a generally more
active star, for which the average Prot is well known (11.73±0.07 days, Croll et al. 2007). This star has relatively strong K emission with hK′i=1.231 and large intranight variability as shown in Figure 11 which varies on time-scales at least as short as the length of the indi- vidual exposures (30 minutes). The average emission from night to night clearly varies with the star’s rotation though with a lower amplitude (0.011 A)˚ as compared to HD 73256 (0.045 A),˚ likely because the star is intrinsically more active, with a larger percentage of its surface covered in spots. This makes it difficult to extract any magnetic and/or tidal contribution to HD 189733’s chromospheric emission by the planet as no correlation is seen between the planet’s orbit and the residuals to the rotational modulation. However, Figure 12 shows how the integrated mean absolute deviation of the K line residuals from the global mean per – 11 – night (MADK) vary with orbital phase. Though we only have four nights of observations spanning 1.4 orbits, there is a clear increase in very-short-term (≤30 minutes) activity at
φorb ∼ 0.8. Remarkably, this is the same phase at which SPI peaks for HD 179949 and τ Boo as measured both spectroscopically and photometrically. (See discussion below and Walker et al. 2006, 2007)
3.3. τ Boo: SPI on a tidally-locked star
The star with the shortest rotation period in our sample is τ Boo. It has the largest vsini (= 14.8 m s−1, Gray 1982) and is believed to be in synchronous rotation with its tightly- orbiting massive planet (Prot = 3.2 ± 0.5 d, Henry et al. 2000, Porb = 3.31250 d, Mpsini=4.4 MJ , Butler et al. 1997). We observed a small but significant night-to-night modulation in the H & K emission of τ Boo during the first 3 years of observations with no obvious phasing with the planet’s position. (Data from 2001, 2002 and 2003 are in Figure 7 of Shkolnik et al. 2005a.) Due to the tidal-locking of the star with the planet, we must depend on consistent orbital phasing of any modulated activity to disentangle SPI from rotational modulation for τ Boo. Walker et al. (2007) presented light curves of τ Boo taken in 2004 and 2005 observed in broadband optical light by the MOST space telescope. In the first year, they observed a significant photometric signal close to the planet’s orbital frequency. In the second year, there was no signal of similar strength but a clear correlation with the MAD of the photometry with orbital phase. They showed that when phased to the planet’s orbital period, the active region precedes the sub-planetary point by 68◦, very close to the phase lead we observe in the enhanced activity on HD 179949 and in the MAD of HD 189733’s Ca II K emission. Though synchronisity with the exact planet’s position is not obvious in τ Boo’s 2001−2003 Ca II data, Walker et al. (see their Figure 6) show that the MAD of these data during the
photometrically active phase range, centered on φorb =0.8, is twice as high than outside of it. Though of smaller amplitude (and larger error bars) than in previous years, our new 2006 2 Ca II data (Figure 13) may also show a weak enhancement between φorb =0.7 and 1.2. If this is indeed the case, it implies that an active region leading the sub-planetary point, has persisted on τ Boo for at least 5 years, equivalent to ≈550 planetary orbits.
2Of the 9 June 2006 nights on which τ Boo was observed, three were taken in ‘spectropolarimetry’ mode, those at orbital phases of 0.19, 0.49 and 0.82. It is interesting to note that on the two nights with high K emission, Catala et al. (2007) detect a clear Stokes V signature while the third night, (φorb=0.49) has both low K emission and no Stoke V signal. – 12 –
3.4. Night-to-night activity correlates with planet’s magnetic moment
S´anchez-Lavega (2004) looked at the internal structure and the convective motions of giant extrasolar planets in order to calculate their dynamo-generated surface magnetism. Given the same angular frequency (which is a reasonable approximation for the short-period planets in question), the magnetic dipole moment, and hence the magnetospheric strength, increases with planetary mass. This is observed for the magnetized planets in our own so- lar system, where the magnetic moment grows proportionally with the mass of the planet 2 −1 (Stevens 2005), and more specifically, with the planet’s angular momentum (L ∝ MpRpPp,rot, Arge et al. 1995). Since only lower limits exist for the masses of most hot Jupiters and at such small semi-major axes they should be tidally locked (Pp,rot = Porb), we plot Mpsini/Porb against hMADKi, the average of the integrated MAD of the K line residuals per observing run, in Figure 14. Though we are able to include only one additional active point (HD 189733) to the original plot of Shkolnik et al. 2005a, we continue to see an intriguing corre- lation between the planet’s magnetic moment and the night-to-night chromospheric activity on its star. Of our sample, τ Boo has the most massive planet and yet falls well below the correlation. This is consistent with the proposed Alfv´en wave model where the near zero relative motion due to the tidal locking of both the star and the planet (P∗,rot = Pp,rot =
Porb) produces minimal SPI because of the weak Alfv´en waves generated as the planet passes through the stellar magnetosphere, thereby transporting little excess energy to the stellar surface along the magnetic field lines (Gu et al. 2005). If this correlation between short-term activity and planetary magnetic moment holds for more hot Jupiter systems engaging in SPI, this could provide an empirical tool with which to estimate the strength of extrasolar planetary magnetic fields.
3.5. Long-term stellar activity cycles
With CFHT data spanning 5 years, we can compare the long-term variations in the chromospheric level of HD 179949, τ Boo, υ And and HD 209458. We measure Ca II K emission strength hKi by integrating across the normalized K cores bounded by the K1 features (Montes et al. 1994) and plot their average for each observing run in Figure 15. The error bars represent the MAD value for each observing season. There does not appear to be any correlation between the mean chromospheric activity and the level of night-to-night modulation, be it due to SPI or stellar rotation. Though 5-year baseline is a good start to tracking the intrinsic stellar activity cycles of these stars, limiting their periods to >10 years, the variability from run to run may also be due to active regions on the visible disk of the star. We require more frequent monitoring over several more years to firmly say anything – 13 – more about the activity cycle of any individual program star.
4. Correlating with other activity indicators
A similar analysis to the Ca II K line described above was performed for the Ca II H, IRT line at 8662A˚ and Hα though with the normalization points set at 0.7 of the local continuum for the latter two. This level is within the photospheric damping wings of the line, focusing the analysis on the chromospheric core and excluding some blended and telluric lines. For both HD 179949 and HD 189733, there is a strong correlation between the residuals of Ca II K with those of the Ca II H and 8662A˚ lines (Figure 16) as expected given the common upper energy level of their transitions. However, a poorer correlation exists with the Hα lines. Though Hα is often demonstrated to be just as good a tracer of chromospheric activity as Ca II, a recent analysis by Cincunegui et al. (2007) has shown that when comparing a sample of stars, the correlation between Hα and Ca II is the result of the correlation of each line with spectral type rather than with stellar activity. When comparing the variability of the lines for an individual star, there is no consistent correlation. This is likely the effect of the differing underlying formation physics between them. (Soderblom et al. 1993, Cram & Mullan 1985). We leave the relative energy emitted in the lines and their implications for SPI models for a later paper. The He I D3 line (a blended triplet) at 5876 A˚ correlates well with plage regions on the solar surface (Landman 1981) as well as with the Ca II H&K emission (Saar et al. 1997). It forms in the upper chromosphere and is thought to be back-heated by the stellar corona giving us a unique optical view of this hot plasma. The absence of He I absorption in non- magnetic regions on the sun and in other inactive stars indicates that He I has no basal (acoustically heated) flux level, unlike the other activity indicators in the visible spectrum, and is therefore purely a signature of magnetic activity. We show the spectral region near the He I lines of our program stars in Figure 17. The line is blended with the weak lines of Fe I 5876.30, Cr I 5876.55, and unidentified lines at 5875.76 and 5875.14 A,˚ both blended with telluric H2O (Moore et al. 1966). The im- perfect removal of the telluric lines from our spectra left residuals at the level of .0.003 of the nearby continuum. This made it difficult to analyze night-to-night variations in the stars that exhibit relatively strong variability in the Ca II lines, though the mere presence of the line indicates magnetic heating source, with strong absorption implying great activ- ity. HD 179949, HD 189733 and τ Boo all have clear He I absorption while HD 209458, HD 149143, HD 217107 as well as our standard star 61 Vir have only weak, if any, absorp- tion. Though night-to-night variability is difficult to quantify in HD 179949 and HD 189733, – 14 – there is a clear increase in He I D3 absorption on the night that each of the two stars displays its maximum Ca II emission, which do not occur on the same night.3 We measured the average He I EW by deblending it with the contaminated lines, a technique particularly difficult for the rapid rotator τ Boo. The values are listed in Table 1. It is interesting to note that the He I EW does not follow the power-law correlation with Ca II H & K emission observed by Saar et al. (1997) for G and K dwarfs, but does display a strong correlation with the short-term activity metric hMADKi plotted in Figure 18, where stars with more night-to-night activity have stronger He I absorption. The relationship between the He I absorption and stellar rotation period observed by Saar et al. is also not obvious in our small sample. It therefore remains possible that the strength of the He I D3 line may predict whether a planet-bearing star will have night-to-night variability. When more stars with hot Jupiters are discovered this may come in handy in helping to decide which systems should be studied further with intensive time-series observations.
5. Summary
We have observed 7 stars with hot Jupiters using CFHT’s ´echelle spectrograph ES- PaDOnS to search for night-to-night modulation of the Ca II emission for evidence of SPI. Four of these have been observed in our previous studies of time-varying Ca II H & K: HD 179949, υ And, HD 209458, and τ Boo, and we have added three new targets: HD 189733, HD 217107 and HD 149143. For our prime target, HD 179949, we now have a total of six observing runs spanning 5 years. During four runs (Aug 2001, July 2002, Aug 2002 and Sept 2005) the Ca II emission varied with the orbital period of 3.092 days, with consistent amplitude and peak phase indicative of a magnetic interaction between the star and planet. The peak activity on HD
179949 in these epochs occurs at φorb ≈ 0.8, leading the sub-planetary longitude by some 70◦. Interestingly, this same phase shift is observed in the MAD of the Ca II K residuals of both τ Boo and HD 189733. The phase lead can provide information on the field geometries (i.e. Parker spiral) and the nature of the effect such as tidal friction, magnetic drag or reconnection with off-center magnetic fields. HD 179949 data from the other two runs (Sept 2003 and June 2006) clearly vary with
3We cannot measure the He I EW of HD 179949 from the September 2005 data since poor weather prevented us from observing a telluric standard. Similarly, a telluric standard was not observed for the spectropolarimetric study of υ And. – 15 – the rotation period of 7 days. A similar effect is seen on υ And where one of four epochs appears to be modulated by rotation rather than the planet’s motion. This on/off behavior has been modeled by Cranmer & Saar (2007) to be an effect of magnetic reconnection with the stellar field as it varies with the star’s long-term activity cycle. We present the expected correlations between the variability observed in the Ca II K line with Ca II H and IRT 8662, and a weaker correlation with Hα. Though we could not accurately measure the variability in the upper-chromosphere line He I D3, we show that it has the potential to flag stars which might be active in Ca II K in on a night-to-night time scale. To date we have observed 13 stars with hot Jupiters at CFHT and VLT, of which 5 appear to be actively engaging in SPI: HD 179949, υ And, HD 189733, HD 73256 and τ Boo. The activity as measured by the mean absolute deviation over a run on the first
four of these stars correlates well with Mpsini/Porb, a value proportional to the planet’s magnetic moment, and thus with the hot Jupiter’s magnetic field strength. Because of their small separation (≤ 0.075 AU), a hot Jupiter lies within the Alfv´en radius of its host star, allowing a direct magnetic interaction with the stellar surface. Although this correlation is tentative, short-term chromospheric variability may be our first probe of extrasolar planetary magnetospheres.
We are very grateful to Claude Catala and John Barnes for contributing several τ Boo and υ And spectra, to Jean-Francois Donati for making Libre Esprit available to CFHT users, to the CFHT staff for their excellent support of this program, and to Benjamin Brown for useful discussions on planetary dynamos. Research funding from the NASA Postdoctoral Program (formerly the NRC Research Associateship) for E.S. and the National Research Council of Canada (D.A.B.) are gratefully acknowledged.
REFERENCES
Arge, C.N., Mullan, D.J., Dolginov, A.Z., 1995, ApJ, 443, 795
Bakos, G. A.,´ Knutson, H., Pont, F., Moutou, C., Charbonneau, D., Shporer, A., Bouchy, F., Everett, M., Hergenrother, C., Latham, D. W., Mayor, M., Mazeh, T., Noyes, R. W., Queloz, D., P´al, A., Udry, S., 2006, ApJ, 650, 1160
Bouchy F., Udry S., Mayor M., Moutou C., Pont F., Iribarne N., Da Silva R., Llovaisky S., Queloz D., Santos N.C., Segransan D., Zucker S., 2005, A&A, 444, L15 – 16 –
Butler, R.P., Marcy G.W., Williams, E., Hauser, H., Shirts, P., 1997, ApJ, 474, L115
Catala, C., Donati, J.-F., Shkolnik, E., Bohlender, D., Alecian, E., 2007, MNRAS, 374, 42
Catalano, S., Rodon`o, M., Frasca, A., Cutispoto, C., 1996, IAUS, 176, 403
Charbonneau, D., Brown, T., Latham, D., Mayor, M., Mazeh, T., 1999, ApJ, 529, 45
Cincunegui, C., D`ıaz, R.F., Mauas, P.J.D., 2007, A&A, 469, 309
Cram, L.E., Mullan, D.J., 1985, ApJ, 294, 626
Cranmer, S. & Saar, S., 2007, arXiv:astro-ph/0702530v1
Croll, B., Matthews, J., Walker, G., 2007, ApJ, submitted
Cumming, A., Marcy, G.W., Butler, R.P., 1999, ApJ, 526, 890
Cuntz, M., Saar, S.H., Musielak, Z.E., 2000, ApJ, 533, L151
De Silva et al. 2006, A&A, 446, 717
Donati, J.-F., Semel, M., Carter, B. D., Rees, D. E., Collier Cameron, A. 1997, MNRAS, 291, 658
Donati, J.-F. 2007, MNRAS, in preparation
Fischer D., Marcy G., Butler P., Vogt S., Apps K., 1998, PASP, 111, 50
George, Samuel, Stevens, Ian, 2007, arXiv:0708.4079v1
Glebocki, R., Bielicz, E., Pastuszka, Z., Sikorski, J., 1986, AcA, 36, 369
Gray, D.F., 1982, ApJ, 255, 200
Griessmeier, J.-M. Stadelmann, A., Penz, T., Lammer, H., Selsis, F., Ribas, I., Guinan, E. F., Motschmann, U., Biernat, H. K. Weiss, W. W., 2004, A&A, 425, 753
Gu, P.-G., Shkolnik, E., Li, S.-L., Liu, X-W, 2005, AN, 326, 909
Henry, G.W., Baliunas, S.L., Donahue, R.A., Fekel, F.C., Soon, W., 2000, ApJ, 531, 415
Ip, W.-H., Kopp, A., Hu, J.-H., 2004, ApJ, 602, L53
Kashyap, V., Drake, J.,Steve Saar, S., 2006, Cool Stars 14 - Abstract # 208 – 17 –
Khodachenko, M. L., Lammer, H., Lichtenegger, H. I. M., Langmayr, D., Erkaev, N. V., Griemeier, J.-M., Leitner, M., Penz, T., Biernat, H. K., Motschmann, U., Rucker, H. O., 2007, P&SS, 55, 631
Landman, D.A., 1981, ApJ, 251, 768
Lazio, T.J.W., Farrell, W.M., 2007, arXiv:0707.1827v1
Mazeh, T., Naef, D., Torres, G., Latham, D.W., Mayor, M., Beuzit, J.-L., Brown, T.M., Buchhave, L., Burnet, M., Carney, B.W., Charbonneau, D., Drukier, G., Laird, J.B., Pepe, F., Perrier, C., Queloz, D., Santos, N.C., Sivan, J.-P., Udry, S., Zucker, S., 2000, ApJ, 532, L55
McIvor, T., Jardine, M., Holzwarth, V., 2006, MNRAS, 367, 1
Montes, D., Fernandez-Figueroa, M.J., de Castro, E., Cornide, M., 1994, A&A, 285, 609
Moutou, C., Donati, J.-F., Savalle, R., Hussain, G., Alecian, E., Bouchy, F., Catala, C., Collier Cameron, A., Udry, S., Vidal-Madjar, A., 2007, A&A, 473, 651
Noyes, R.W., Hartmann, L.W., Baliunas, S.L., Duncan, D.K., Baughan, A.H., 1984, ApJ, 279, 763
Olson, P., Christensen, U. R., 2006, E&PSL. 250, 561
Pont, F., Bouchy, F., Melo, C., Santos, N. C., Mayor, M., Queloz, D., Udry, S., 2005, A&A, 438, 1123
Preusse, S., Kopp, A., B¨uchner, J., Motschmann, U. 2006, A&A, 460, 317
Saar, S. H., Huovelin, J., Osten, R. A., Shcherbakov, A. G. 1997, A&A 326, 741
Saar, S.H., Cuntz, M., Shkolnik, E., 2004, in IAU Symp. 219, Stars as Suns: Activity, Evolution, Planets, eds. A.K. Dupree and A.O. Benz, p.355
Saar, S., Kashyap, V., Cuntz, M., Shkolnik, E., Hall, J., 2006 Cool Stars 14 Meeting - Abstract # 252
S´anchez-Lavega, A., 2004, ApJ, 609, 87
Shkolnik, E., Walker, G.A.H., Bohlender, D.A., 2003, ApJ, 597, 1092
Shkolnik, E., Walker, G.A.H., Bohlender, D.A.,Gu, P.-G., K¨urster, M., 2005a, ApJ, 622, 1075 – 18 –
Shkolnik, E., Walker, G.A.H., Rucinski, S., Bohlender, D.A., Davidge, T.J., 2005b, AJ, 130, 799
Soderblom, D.R., Stauffer, J.R., Hudon, J.D., Jones, B.F., 1993, ApJS, 85, 315
Stevens, I. R. 2005, MNRAS, 356, 1053
Tinney, C., Butler, P., Marcy, G., Jones, H., Penny, A., Vogt, S., Apps, K., Henry, C., 2001, ApJ, 551, L507
Udry, S., Mayor, M., Clausen, J.V., Freyhammer, L.M., Helt, B.E., Lovis, C., Naef, D., Olsen, E.H., Pepe, F., Queloz, D., Santos, N.C., 2003, A&A, 407, 679
Valenti, J. A., Fischer, D. A., 2005, ApJS, 159, 141
Vidal-Madjar, A., Lecavelier des Etangs,´ A., D´esert, J.-M., Ballester, G., Ferlet, R., H´ebrard, G., Mayor, M., 2003, Nature, 422, 143
Vidal-Madjar, A., D´esert, J.-M., Lecavelier des Etangs, A., H´ebrard, G., Ballester, G. E., Ehrenreich, D., Ferlet, R., McConnell, J. C., Mayor, M., Parkinson, C. D., 2004, ApJ, 604, L69
Walker, G.A.H., Matthews, J.M., Kuschnig, R., Johnson, R., Rucinski, S., Pazder, J., Burley, G., Walker, A., Skaret, K., Zee, R., Grocott, S., Carroll, K., Sinclair, P., Sturgeon, D., Harron, J., 2003b, PASP, 115, 1023
Walker G. A. H., Matthews, J.M., Kuschnig, R., Rowe, J.F., Guenther, D. B., Moffat, A. F. J., Rucinski, S.M., Sasselov, D., Seager, S.,Shkolnik, E., Weiss, W. W. 2006, in Arnold L., Bouchy F., Moutou C., eds, Proc. Haute-Provence Observatory Coll., Tenth An- niversary of 51 Peg- B: Status of and Prospect for Hot Jupiter Studies. Frontier Group, Paris, p. 267 (http://www.obs-hp.fr/www/pubs/Coll51Peg/proceedings.html)
Walker, G.A.H, Croll, B., Matthews, J.M., Kuschnig, R., Huber, D., Weiss, W.W., Shkolnik, Rucinski, S.M., Guenther, D.B., Moffat, A.F.J., Sasselov, D., 2007, ApJ, submitted
Wolf, M., Harmanec, P., 2004, IBVS, 5575, 1
Wright, J., Marcy, G., Butler, R., Vogt, S., 2004, ApJS, 152, 261
Zarka, P., Treumann, R.A., Ryabov, B.P., Ryabov, V.B., 2001, Ap&SS, 277, 293
This preprint was prepared with the AAS LATEX macros v5.2. – 19 –
1 1 Ca II K Ca II H 0.8
0.6 0.9 0.4 He I
Normalized Flux 0.2
0 0.8 3920 3940 3960 3980 5872 5874 5876 5878
1 1
0.8 0.8
0.6 0.6
Normalized Flux 0.4 Hα Ca II IRT 0.4 0.2 6550 6560 6570 8450 8500 8550 8600 8650 8700 Wavelength (Å) Wavelength (Å)
Fig. 1.— Selected regions of a normalized spectrum of HD 189733 identifying key stellar activity diagnostics. – 20 –
HD 217107 HD 209458 1
0.8
0.6 Normalized Flux 0.4
0.2
HD 149143 τ Boo 1
0.8
0.6 Normalized Flux 0.4
0.2 3932 3934 3936 3932 3934 3936 Wavelength (Å) Wavelength (Å)
Fig. 2.— The mean normalized Ca II K cores for four of the six program stars. That of HD 179949, HD 189733 and υ And are shown in Figures 3, 4 and 5, respectively. – 21 –
September 2005 June 2006
1
0.8
0.6 Normalized Flux 0.4 0.02
0.01
0
Residual Flux -0.01
-0.02 0.015
0.01 MAD 0.005
0 3932 3934 3936 3932 3934 3936 Wavelength (Å) Wavelength (Å)
Fig. 3.— Top: The mean normalized Ca II K emission of HD 179949 on the three nights observed in September 2005 and the four nights in June 2006. Middle: The residuals relative to their respective means. Bottom: The mean absolute deviation (MAD) of the residuals. – 22 –
3
2.5
2
1.5
1
Normalized Flux 0.5
0
0.02
0.01
0
-0.01 Residual Flux -0.02
0.025
0.02
0.015
MAD 0.01
0.005
0 3932 3933 3934 3935 Wavelength (Å)
Fig. 4.— As in Figure 3 for HD 189733 observed in June 2006. – 23 –
0.8
0.6
0.4 Normalized Flux 0.2
0.005
0
Residual Flux -0.005
0.006
0.004 MAD
0.002
0 3931 3932 3933 3934 3935 3936 Wavelength (Å)
Fig. 5.— As in Figure 3 for υ And observed in September 2005. – 24 –
0.03 2001, 2002 2005 0.025
0.02
0.015
0.01 Integrated Residual Ca II K Flux
0.005
0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 φ orb
Fig. 6.— Integrated flux of the K-line residuals from a normalized mean spectrum of HD 179949 as a function of orbital phase for the 2001 and 2002 data (open circles) pub- lished in Shkolnik et al. (2003) and 2005 data (filled circles). The grey line is a best-fit spot model to the earlier data whose thickness reflects the error in the phase shift. The black line is the same fit slightly shifted in phase by -0.07 to better fit the 2005 data. This small shift relative to the earlier data is not signficant. Error bars in the integrated residual K flux are ± the intranight residual RMS. Note that the data points are repeated for two cycles. – 25 –
0.03 2003 2006 0.025
0.02
0.015
0.01 Integrated Residual Ca II K Flux
0.005
0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 φ orb
Fig. 7.— Integrated flux of the K-line residuals from a normalized mean spectrum of HD 179949 for 2003 and 2006 data plotted on the 3.092-day orbital period with the SPI spot model from Figure 6 over-plotted. Error bars are ± the intranight residual RMS. – 26 –
0.02 2006 2003
0.015
0.01
Integrated Residual Ca II K Flux 0.005
0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 φ rot
Fig. 8.— Integrated flux of the K-line residuals from a normalized mean spectrum of HD 179949 for 2003 and 2006 data plotted on a 7-day rotation period with phases rela- tive to the first night of each run. The points are vertically shifted such that the minimum of each curves is zero. Error bars are ± the intranight residual RMS. The curves are best-fit spot models to the two data sets. – 27 –
0.004
0.002
0
-0.002 Integrated Residual Ca II K Flux -0.004
-0.006 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 φ orb
Fig. 9.— Integrated flux of the K-line residuals from a normalized mean spectrum of υ And for September 2005 data plotted on the 4.6-day orbital period. Error bars are within the size of the of the points. The curve is the best-fit spot model to our 2002 and 2003 data. (See Figure 8 of Shkolnik et al. 2005.) – 28 –
0.006
0.004
0.002
0
-0.002 Integrated Residual Ca II K Flux -0.004
-0.006
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 φ rot
Fig. 10.— Integrated flux of the K-line residuals from a normalized mean spectrum of υ And for September 2005 data plotted on a 12-day rotation period. Error bars are within the size of the of the points. The curve is a best-fit spot model. – 29 –
0.03
0.02
0.01
0
-0.01
-0.02
-0.03
Integrated Residual Ca II K Flux -0.04
-0.05
-0.06 0.5 0.6 0.7 0.8 0.9 1 φ rot
Fig. 11.— Integrated flux of the K-line residuals from a normalized mean spectrum of HD 189733 as a function of a 11.7-day rotational phase (Moutou et al. 2007, Croll et al. 2007). The open circles are residuals of individual exposures relative to a global mean, while the solid circles use only the mean of each night. Error bars represent twice the measurement error of a given exposure as measured by the residuals outside of the Ca II K core. The solid curve is a best-fit sinusoid tracing the rotation of the star. – 30 –
0.03
0.025
0.02
0.015
0.01 MAD of Nightly Ca II K Residuals 0.005
0 0 0.5 1 1.5 2 φ orb
Fig. 12.— Mean absolute deviation (MAD) of the nightly K-line residuals shown in Figure 11 of HD 189733 as a function of the 2.2-day orbital period. The error bars represent the integrated MAD immediately outside the Ca II emission core and reflect the S/N obtained for each night. Errors in phase are less than the size of the points. – 31 –
0.004
0.003
0.002
0.001
0
-0.001
Integrated Residual Ca II K Flux -0.002
-0.003
-0.004 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 φ orb
Fig. 13.— Integrated flux of the K-line residuals from a normalized mean spectrum of τ Boo as a function of orbital phase with a best-fit spot model. Note that since Prot ≃ Porb for this system, φrot ≃ φorb. The three observations taken in ‘spectropolarimetry’ mode are at φrot = 0.19, 0.49 and 0.82. Error bars are ± the intranight residual RMS. – 32 –
0.006
0.005
0.004
0.003
0.002 τ Boo
0.001
0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Mpsini/Porb
Fig. 14.— The ratio of the minimum planetary mass (in Jupiter masses) to the orbital period (in days) plotted against the average MAD of the K-line per observing run for all 13 stars observed. The x-axis quantifies the planet’s magnetic moment assuming tidal locking, such
that Prot = Porb. The filled-in circles are of stars which exhibit significant night-to-night variability in the Ca II K line: HD 73526, υ And, HD 179949 and HD 189733 (this work). The tidally-locked system of τ Boo does not follow the correlation traced out by the others. The error bars are one-sided due to the positive contribution of integrated MAD immediately outside the Ca II emission core and reflect the S/N obtained for each target. – 33 –
0.4 HD 179949 0.39
0.38
0.37
0.36
0.35 τ Boo 0.34
0.33
0.32
0.31 υ And 0.26
0.255
Mean Integrated K emission 0.25
0.245
0.21 0.205 HD 209458 0.2 0.195 0.19 0.185 0.18 2001 2002 2003 2004 2005 2006 2007 Year
Fig. 15.— The average integrated Ca II K emission of HD 179949, τ Boo, υ And and HD 209458 for each of the five or six observing runs. The error bars are the MAD values for each observing season. – 34 –
HD 179949 HD 189733 0.02 0.01 0.01 0
0 Ca II H -0.01
0.99 0.94 -0.02 -0.01
0.005 0.004
0 0.002 8662 λ 0 -0.005 Ca II 0.98 -0.002 0.95 -0.01 0.005 0.02
0 0 α
H -0.02 -0.005 -0.04 0.89 0.86
-0.06 -0.01 -0.02 -0.01 0 0.01 -0.01 0 0.01 0.02 Integrated Flux in Ca II K Residuals Integrated Flux in Ca II K Residuals
Fig. 16.— Integrated flux in the Ca II K residuals plotted against the residual flux in Ca II H, the 8662A˚ line and Hα for HD 179949 and HD 189733. The number in the corner is the correlation coefficient of the best-fit line. – 35 –
1
0.95
0.9
0.85 Normalized Flux τ Boo HD 179949 1
0.95
0.9
0.85 Normalized Flux HD 209458 HD 217107 1
0.95
0.9
0.85 Normalized Flux HD 149143 61 Vir
5872 5874 5876 5878 5872 5874 5876 5878 Wavelength (Å) Wavelength (Å)
Fig. 17.— Normalized spectra in the region of the He I line (5876 A)˚ for 5 program stars plus the standard, 61 Vir. The vertical line corresponds to the location of the line. The same spectral region for HD 189733 is shown in Figure 1. The He I cannot be measured for υ And since no telluric standard was observed on those nights. – 36 –
0.005
0.004
0.003
0.001
0 0 10 20 30 40 He I EW (mÅ)
Fig. 18.— Average MAD of the K emission hMADKi for the 2006 program stars as a function of the He I equivalent width. The error bars reflect the difficulty in measuring the EW due to line blends, especially for the rapidly rotating τ Boo.